1. Field of the Invention
The present invention relates to a computer program product, system, and method for recovering from a pending uncompleted reorganization of a data set.
2. Description of the Related Art
In certain computing environments, multiple host systems may configure data sets in volumes configured in a storage system, such as interconnected storage devices, e.g., a Direct Access Storage Device (DASD), Redundant Array of Independent Disks (RAID), Just a Bunch of Disks (JBOD), etc. Data sets are comprised of extents, which may comprise any grouping of tracks and data storage units. The Z/OS® operating system from International Business Machines Corporation (“IBM”) has a Volume Table of Contents (VTOC) to provide information on data sets of extents configured in the volume, where the VTOC indicates the location of tracks, extents, and data sets for a volume in storage. For Virtual Storage Access Method (VSAM) data sets, the z/OS operating system provides VSAM volume data sets (VVDS) stored in a data set to provide information on VSM key sequenced data sets. (Z/OS is a registered trademark of IBM in the United States and other countries)
Data sets may be reorganized to improve the access performance of the data set when over time the data or records in the data set are rearranged to be in a less than optimal access state. Data reorganization may order the records or data in the data set according to an index key or other ordering to improve user access performance. When proper procedures are not followed during the data reorganization, data integrity problems may result in outages and data loss. For instance, during reorganization, no other client is supposed to access the data set and modify the data. However, not all clients will follow serialization, which may result in corrupted data. If a reorganization fails, then applications will have to continue operating on the data set whose data is not arranged in an ordering that optimizes access.